Log Road models logging roads based on existing OSM street data, forest stands, and a cost surface.

####Input

OSM street data	Forest Stands	Cost Surface
OSM	Shapefile (EPSG3857)	GeoTIFF (EPSG32610)

####Process

• Buffer from the stands are created based on a specified skidding distance.
• Fishnet grid covering the buffer extent is created.
• Dictionary is created. For each grid cell the intersecting buffers are recorderd: {GridCellID:[BufferID1,BufferID2,...]}
• OSM data are downloaded and get burned into the Cost Surface
• Cost Surface gets updated: Areas outside of the stands are more expensive, areas at the boundary of the stands are less expensive
• Buffers intersecting with OSM streets are removed from the dicitionary.
• Loop over Dictionary:
  • Cell from dictionary with most intersecting buffers is selected.
  • New road from cell to closest road is created.
  • Buffers intersecting with new road are removed from the dicitionary.
• New roads are written to output shapefile.

####Output

New Roads
Shapefile (EPSG3857)

Requires python-gdal, skimage.graph, requests, and numpy

To install, simply python setup.py install or work directly from the root directory.

Ideally the CostSurface consists for actuall $ US values in order to estimate the actual costs of construction to roads. But for the porpuse to identify the length and location of the new roads, a relativ coast surface is sufficient.

test_costsurface.py with CostSurface.py can be used to create a cost surface based on variables slope and river (buffered rivers). In CostSurface.py the relation between the two can be specified. We received the best test results with the following parameters:

    costSurfaceArray[costSurfaceArray > 50] **=2
    costSurfaceArray[riverArray == 1] **=5